آثار ریزساختار، اسپرد قیمت خرید و فروش و نوسانات بازار ارز لحظه ای در قبل و پس از EMU

This article examines how microstructure effects, evident in high frequency data, influence bid–ask spreads and volatility in transaction price series. It uses the event of European Monetary Union (EMU), and the upheaval that this entailed, as an opportunity to empirically investigate these relationships in the electronic inter-dealer spot FX market. The microstructure effects relate to both price and time. There are two price effects, namely price discreteness and price clustering, and two time effects, namely the time elapsed between sample periods and the time-gap between successive trades or quoted price submissions. Strong evidence emerges that all four factors are important in the determination of bid–ask spreads. This study uses a unique and rich foreign exchange (FX) dataset of global inter-dealer electronic transactions to examine microstructural effects in the spot foreign exchange market. This dataset enables us to shed new light on the debate surrounding the observations that trading volumes have fallen and bid–ask spreads have widened in inter-dealer spot FX markets following European Monetary Union (EMU). Our work provides a more detailed account of the changes that actually occurred at this time, because our data is more comprehensive than has previously been available. Our four-technical-feature explanation is in contrast to the hypothesis of market maker response to exogenous changes in volume as proposed by Hau, Killeen and Moore [Hau, H., Killeen, W., Moore, M. (2000). The euro as an international currency: Explaining puzzling first evidence. Centre for Economic Policy Research, working paper., Hau, H., Killeen, W, Moore, M. (2002). How has the euro changed the foreign exchange market? Economic Policy 17, 34, 149-191]. Price discreteness means that prices or exchange rates are not an infinite number of digits long, but rather they are truncated to a small number of digits. In the case of the FX market, exchange rates are specified to five digit accuracy. Price clustering refers to the fact that traders may not use all available exchange rates uniformly. In practice, rates ending in 0 or 5 tend to be used more than other rates. The time elapsed between the sample periods is important for a very obvious reason— price levels can differ radically if data is sampled from periods that are far apart in time. On the other hand, the time-gap between successive individual prices is also important because it allows these prices to drift apart. When the successive prices are transaction prices, this effect increases volatility. When they are successive bid and ask prices, the bid–ask spread is increased. EMU brought widespread change to financial markets. Much of this change is directly due to the re-denomination of certain instruments from Deutschemarks (DEM) to euros (EUR). Since these currency units are of different values, the nature of the price discreteness affecting instruments which are now denominated in EUR will be different from what it was under DEM denomination. This point is exemplified by our finding that the smallest sized bid–ask spread and smallest price increment for the EUR are both 74% greater than that for the DEM, after controlling for drift in currency values. Our four proposed factors are successful in explaining the observed changes in bid–ask spreads, but are less able to explain the observed changes in price volatility. Also, our results overwhelmingly accept the price resolution hypothesis explanation for price clustering behavior in the spot FX market and overwhelmingly reject the price attraction hypothesis. In the case of the EUR(DEM)/USD bid–ask spread, we provide a deeper understanding of the technical market features that caused this to increase. We show that the widening of the USD/JPY bid–ask spread seems primarily due to the inter-temporal change in currency value. We also show that the narrowing of EUR(DEM)/CHF bid–ask spreads seems largely due a near 50% decrease in the pricing increment used. We find that increased volume has reduced the time-gap for traded and quoted prices for USD/CHF. Finally, in the case of EUR(DEM)/JPY, we find that market practice caused wider bid–ask spreads. The bid–ask spread data evidence suggests that the advent of EMU seems to have strengthened the USD's position as the dominant international vehicle currency. We consider this surprising because we believe that part of the intention in launching the single currency must surely have been the opposite.

This study uses a unique and rich foreign exchange (FX) dataset of global inter-dealer electronic transactions to examine microstructural effects in the spot foreign exchange market. This dataset allows us to shed new light on the debate surrounding the observations that trading volumes have fallen and bid–ask spreads have widened in inter-dealer spot FX markets following European Monetary Union (EMU). Our work provides a more detailed account of the changes that actually occurred at this time, because our data is more comprehensive than has previously been available. However, this is more than just an empirical study. We also introduce new theoretical explanations and new methodologies to this debate. These enable us to demonstrate that much of the change in bid–ask spreads and some of the change in price volatility can be explained by four technical features of high-frequency data, namely price discreteness, price clustering, the time elapsed between sample periods and the time-gap between successive traded or quoted prices. This explanation is in contrast to the hypothesis of market maker response to exogenous changes in volume proposed by Hau et al., 2000 and Hau et al., 2002. It is widely acknowledged that volumes decreased in the inter-bank spot FX market after EMU (see BIS, 2001). It also became increasingly accepted that the EUR/USD bid–ask spread widened at the same time. Hau et al., 2000 and Hau et al., 2002 suggested that lower FX trading volumes and wider bid–ask spreads observed since EMU, are both due to a decrease in “market transparency”. The latter hypothesis centers on the idea that the availability of fewer currency pairs after EMU makes risk management harder to implement. Hau et al. suggest that this causes market makers to quote wider bid–ask spreads, which in turn results in lower volumes. One part of our argument has been made previously by Goodhart, Love, Payne, and Rime (2002). They argued that “price granularity” could account for the observed fall in inter-dealer bid–ask spreads and that reduced volumes are a coincidence due to unrelated structural changes in the industry. What they call “price granularity” is more usually called “price discreteness” in the literature. Price discreteness means that prices are not an infinite number of digits long, but rather they are truncated to a small number of digits. In the case of the FX market, exchange rates are specified to five digit accuracy by convention. The Goodhart et al. (2002) view was also echoed by Detken and Hartmann (2002) who used lower frequency data. We believe that Goodhart et al. (2002) were broadly on the right track, but they did not uncover the full story. In addition, Goodhart et al. (2002) only focused on the USD/DEM and EUR/USD exchange rates. We are able to analyze many more currency pairs over the period of EMU convergence. The price clustering and time arguments that we present are new to this debate. Price clustering refers to the fact that traders may not use all available exchange rates uniformly. In practice, rates ending in 0 or 5 tend to be used more than other rates. The time elapsed between the sample periods is important for a very obvious reason— price levels can differ radically if data is sampled from periods that are far apart in time. On the other hand, the time-gap between successive individual prices is also important because it allows these prices to drift apart. When the successive prices are transaction prices, this effect increases volatility. When they are successive bid and ask prices, the bid–ask spread is increased. These new perspectives reveal that volumes and bid–ask spreads are not unrelated as Goodhart et al. (2002) had argued, but are in fact inextricably interwoven. However, our arguments suggest the opposite causality to that proposed by Hau et al., 2000 and Hau et al., 2002. In short, we find that smaller volumes cause larger bid–ask spreads for technical reasons to do with measurement, whereas Hau et al., 2000 and Hau et al., 2002 had argued that larger bid–ask spreads had caused smaller volumes due to trader behavior. The remainder of this paper is organized as follows. Section 2 discusses the importance of price discreteness and price clustering and reviews that literature. Section 3 similarly explores the role of time and duration. Section 4 reviews market structure and our unique dataset, while Section 5 discusses the empirical methodology. Section 6 presents the results and Section 7 concludes.

This article set out to document and explain the changes in bid–ask spreads and in volatility in five currency-pairs between pre-EMU and post-EMU periods. The data used shows that volatility fell for all exchange rates, while some bid–ask spreads widened and others narrowed. Four distinct potential explanatory market microstructure factors for these changes were explicitly considered: price discreteness; price clustering; the inter-temporal shift in currency value; and the time-gap between successive observations. The four proposed factors are successful in explaining the observed changes in bid–ask spreads, but are less able to explain the observed changes in price volatility. The combination and influence of each of these four factors varies from exchange rate to exchange rate. In characterizing the source of price clustering behavior in the spot FX market, our results overwhelmingly affirm the price resolution hypothesis explanation and firmly reject the price attraction hypothesis. Our price clustering analysis makes use of new test statistics which we propose to capture behaviors described by Goodhart and Curcio (1991). In the specific case of the EUR(DEM)/USD bid–ask spread, we broadly concur with the price granularity (price discreteness) hypothesis of Goodhart et al. (2002) which asserts that redenomination is the most likely cause of an increase in the spread for this exchange rate. We proceed to show that rigid market quoting practices are critical in this. The switch from USD/DEM to EUR/USD quoting practice caused a 74% increase in the one-tick EUR(DEM)/USD bid–ask spread. If the market had been willing to embrace an additional half point price increment by using 0 and 5 in the sixth digit position, the one-pip EUR/USD bid–ask spread would have been 13% lower than for USD/DEM. The fall observed in EUR/USD price clustering had a dissipating effect on this. So too, did the inter-temporal shift in currency value. On the other hand, the increase in the time-gap between prevailing bid and ask prices exacerbated the rise. However, as we demonstrated above, it was the inversion of this quoting format (i.e. using EUR/USD rather than USD/EUR) that was primarily responsible for this bid–ask spread increase. The widening of the USD/JPY bid–ask spread seems primarily due to the sharp inter-temporal change in currency value. Price discreteness can explain no part of this change because there was no re-denomination of the USD/JPY from pre-EMU to post-EMU periods. Like the EUR(DEM)/USD, the fall in price clustering should have had a dampening effect on the bid–ask spread. On the other hand, the increased time-gap between bid and ask prices should contribute to the widening of post-EMU spreads. Finally, wider EUR/JPY and EUR/USD bid–ask spreads preclude any possibility that the EUR could act as a vehicle for currency flows between USD and JPY. This means, at the very least, that the threat of indirect trading which previously may have inhibited the widening of the direct USD/JPY bid–ask spread became a less potent influence post-EMU. The narrowing of EUR(DEM)/CHF bid–ask spreads seems largely due a near 50% decrease in the pricing increment used. Although there is almost no change in the economic value of the usable minimum tick size, implying no change in price discreteness, the price clustering patterns show that the pre-EMU half-points were rarely used in practice. The data show that whole numbers were mainly used and that the price increment used in practice was actually two pips. By contrast, in the post-EMU, EUR denominated world, traders used one pip increments, effectively halving the bid–ask spread. Exchange rate value levels and price clustering patterns over the range of whole numbers were little changed. The large increase in the time-gap will have tempered the narrowing of this bid–ask spread. This explanations tells us that the bid–ask spread change is due to a change in trader behavior but it does not tell us why this change occurred. The rejection of the attraction hypothesis for price clustering suggests that human fondness for round numbers is not a satisfactory explanation. A more plausible reason may be increased competition from USD/CHF. The trend-bucking increase in USD/CHF trade volume closely matches the volume lost by EUR/CHF. If it is the case that USD/CHF has taken over much of the international vehicle currency role of the DEM/CHF, the EUR/CHF could face even more loss of volume if its prices are uncompetitive. Increased volume reduced the time-gap for traded and quoted prices for USD/CHF. Since there is no change in price discreteness and very little change in the currency value level, it appears that the change in the time-gap is the leading contender to explain the narrower bid–ask spreads evident in this exchange rate. The price clustering data also reveals a spread-reducing influence in that it shows less use of extreme 0 and 5 final digit values. It seems plausible that reduced price clustering reflects more accurate pricing behavior in response to increased USD/CHF volume and to the development of a competitive scenario with EUR/CHF. Note that this is the only exchange rate where we find any evidence supporting the Hau et al., 2000 and Hau et al., 2002 assertion that volume changes drive bid–ask spread changes. However, even then, the direction of the volume impetus and the specific exchange rate remain different. Our analysis suggests that higher volume lead to lower bid–ask spreads for USD/JPY, whereas they concluded that lower volume led to higher bid–ask spreads in EUR/USD. In addition, our argument also relies on a competitive vehicle currency scenario. The fact that CHF/JPY was never a liquid currency pair in the inter-dealer market combined with the fact that the CHF has never been considered as a credible international vehicle currency implies that indirect exchange rates are not a significant determinant of either EUR/CHF or USD/CHF bid–ask spread changes. Reconciliation of the evidence on the EUR(DEM)/JPY bid–ask spread is perhaps the most complicated. This spread widened by 58% in the period studied. The value of this exchange rate dropped by 24%, which should indeed lead to a wider bid–ask spread, if the minimum tick change remained unchanged. The value of the minimum tick size did indeed remain almost totally unchanged because the increase in currency unit size from DEM to EUR is almost exactly offset by the decrease in economic value of the minimum pip increment in use post-EMU. This means that the change in exchange rate level between the two sample periods accounts for around half of the observed bid–ask spread increase. An increase in the time-gap and price clustering behavior also contribute to the widening of this bid–ask spread, the size of which more than doubled in pip terms after EMU. The time-gap between successive prices almost triples after EMU, which should also widen bid–ask spreads, due to drift. In addition, the price clustering data reveal that extreme 0 and 5 final digit prices are much more pronounced after EMU, suggesting that the market's quoting practice results in wider bid–ask spreads. We must consider why market practice should have changed in this way. One possible reason is that the EUR/JPY may have lost considerable ground to the USD/JPY as an international vehicle currency. The respective bid–ask spread levels clearly show that there is a substantial saving to be made after EMU by trading EUR/JPY indirectly, using the USD as a vehicle. This provides the economic grounds for indirect trading to flourish. In addition, the steeper decline in EUR/JPY trading volume compared with USD/JPY, suggests that a shift from direct to indirect trading has taken place. This evidence of increased indirect trading from EUR/JPY and the two CHF rates leads to the startling conclusion that the advent of EMU seems to have strengthened the USD's position as the dominant international vehicle currency. Part of the intention in launching the single European currency must surely have been the opposite. The ubiquitous fall in volatility that we observe is more puzzling, not least because it differs so much from other researchers' observations. We conclude that volatility in FX markets is intrinsically a more complicated phenomenon than bid–ask spreads. Volatility may be influenced by information coming to the market at specific times and by temporary imbalances between supply and demand which may differ greatly between data samples. Furthermore, cross currency arbitrage could dissipate across the market what should be localized shocks. Such issues are beyond the scope of this paper and remain for future research.